Determining biodegradation kinetics of ... - ACS Publications

Subscriber access provided by READING UNIV

Article

Determining biodegradation kinetics of hydrocarbons at low concentrations – covering 5 and 9 orders of magnitude of Kow and Kaw Heidi Birch, Rikke Hammershøj, and Philipp Mayer Environ. Sci. Technol., Just Accepted Manuscript • Publication Date (Web): 29 Jan 2018 Downloaded from http://pubs.acs.org on January 29, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32

Environmental Science & Technology

1

Determining biodegradation kinetics of

2

hydrocarbons at low concentrations – covering 5 and

3

9 orders of magnitude of Kow and Kaw

4

Heidi Birch*, Rikke Hammershøj, Philipp Mayer

5

Technical University of Denmark, Department of Environmental Engineering, Bygningstorvet,

6

Building 115, 2800 Kgs. Lyngby, Denmark.

7

ACS Paragon Plus Environment

1


Page 2 of 32

8

Abstract. A partitioning based experimental platform was developed and applied to determine

9

primary biodegradation kinetics of 53 hydrocarbons at ng/L - µg/L concentrations covering C8-

10

C20, 11 structural classes and several orders of magnitude in hydrophobicity and volatility: (1)

11

Passive dosing from a loaded silicone donor was used to set the concentration of each

12

hydrocarbon in mixture stock solutions, (2) these solutions were combined with environmental

13

water samples in gas tight auto sampler vials for 1 to 100 days incubation and (3) automated

14

Solid Phase MicroExtraction (SPME) coupled to GC-MS was applied directly on these test

15

systems for measuring primary biodegradation relative to abiotic controls. First order

16

biodegradation kinetics were obtained for 40 hydrocarbons in activated sludge filtrate, 18 in

17

seawater and 21 in lake water. Water phase half-lives in seawater and lake water were poorly

18

related to hydrophobicity and volatility but were, with a few exceptions, within a factor of 10 or

19

shorter

20

pentamethyldecalin,

21

heptamethylnonane showed limited or inconsistent degradation in all three environmental media.

22

This biodegradation approach can cover a large chemical space at low substrate concentrations,

23

which makes it highly suited for optimizing predictive models for environmental biodegradation.

than

BioHCwin

predictions.

perhydropyrene,

The

most

persistent

hydrocarbons,

1,2,3,6,7,8-hexahydropyrene

and

1,1,4,4,6-

2,2,4,4,6,8,8-

24


2

Page 3 of 32


25

Introduction

26

Biodegradation is the most important removal process for many organic chemicals in the

27

environment. The potential for biodegradation of a chemical structure (biodegradability), or lack

28

thereof, is therefore a key element in the regulatory framework for chemical risk assessment in

29

e.g. Europe and the United States.1,2 The first tier of regulatory biodegradability testing consists

30

of a qualitative screening test for ready biodegradability.3 If chemicals fail this test, higher tier

31

simulation tests can be used to obtain biodegradation half-lives under more realistic

32

environmental conditions including low test concentrations and using environmentally native

33

microorganisms for the test. These simulation tests are expensive and data are much scarcer than

34

from the screening tests.4

35

While qualitative ready biodegradability test data are appropriate for screening of chemicals,

36

environmental biodegradation kinetics are necessary for environmental fate modelling and risk

37

assessment. Different schemes have been proposed to assign biodegradation half-lives to

38

chemicals based on screening test results,4 and models have been developed to predict

39

biodegradability based on chemical structure.5 However, the experimental generation of high

40

quality kinetic data remains crucial, since experimental data remain the gold standard in research

41

and regulation, are the basis for building and refining predictive models and are the ultimate

42

reference for testing such models.5–7 There is a need for larger data sets since comparability in

43

both the test conditions and inoculum is important for the training set data in order to build a

44

model with structural generality.5

45

Risk assessment of petroleum products is complicated as these are complex mixtures of

46

varying composition containing thousands of components, each with their own physicochemical

47

and degradation properties.8 Chemical constituents can be classified according to the chemical


3


Page 4 of 32

48

space either by physical-chemical properties such as air-water partitioning and octanol-water

49

partitioning or by structural grouping and carbon number classes.8–10 The constituents of such

50

products are released to the environment as a mixture, and will therefore be subject to

51

degradation as a mixture. A biodegradation model, BioHCwin, has been developed to predict

52

environmental half-lives for hydrocarbons.6 Although field and grab sample tests in water,

53

sediment and soil were preferred for building the BioHCwin model, screening test data were

54

included for some chemical groups due to lack of data.6 A number of studies have since looked

55

at biodegradation of petroleum product constituents or mixtures at concentrations close to the

56

aqueous solubility of the mixtures or above solubility (dispersions).11–16 Prosser et al.16 compared

57

model predictions to the results from these studies and found the model to perform acceptably as

58

a screening tool. While high substrate concentrations are very relevant for oil spill situations in

59

the proximity of the spill,18 assessing and predicting biodegradation and persistence in the

60

aquatic environment requires data that are obtained at much lower concentrations19. Another

61

study has calculated in situ degradation rates for hydrocarbons based on data from the Deepwater

62

Horizon oil spill, however did not compare the findings to model predictions.17

63

Recently, a new partitioning based experimental platform was introduced for determining

64

biodegradation kinetics of composed mixtures of hydrophobic organic chemicals at

65

environmentally relevant low concentrations.20 Important features of this experimental platform

66

are: (1) Passive dosing from a pre-loaded silicone was used to set concentrations of each

67

constituent in an aqueous mixture without the addition of a co-solvent.21–23 (2) These solutions

68

were combined with environmental water samples containing native microorganisms for

69

incubation in gas tight auto sampler vials for 1 to 100 days. (3) Finally, automated Head Space

70

Solid Phase MicroExtraction (HS-SPME) coupled to GC-MS was applied directly on the test


4

Page 5 of 32


71

systems for determining primary biodegradation kinetics relative to abiotic controls. This

72

approach has recently been applied to a mixture of 9 hydrocarbons with low melting points and

73

high air to water partition ratios,20 and was used to study the effect of inoculum origin on

74

biodegradation kinetics of the 9 hydrocarbons.24

75

The first aim of the present study was to further develop the applicability domain of this

76

experimental platform in order to facilitate biodegradation kinetic testing of larger mixtures of

77

hydrocarbons covering a much wider chemical space. For this purpose the passive dosing

78

technique had to be extended to include different loading principles that in combination are

79

applicable to chemicals with a wider melting point, hydrophobicity, volatility and chemical class

80

range. Solid Phase Microextraction (SPME) had to be operated not only in headspace but also

81

direct submersion mode in order to extend the applicability domain towards the less volatile

82

chemicals and particularly the polyaromatic hydrocarbons (PAHs). The second aim was to apply

83

this new approach for generating a large set of biodegradation kinetic data at environmentally

84

relevant concentrations, covering a substantial part of the chemical space of petroleum

85

hydrocarbons. 53 hydrocarbons were chosen based on (1) ensuring a good coverage of 11

86

structural classes and carbon numbers from 8 to 20, (2) covering a wide range in hydrophobicity

87

and air-water partitioning, (3) our present capabilities with passive dosing, SPME and GC-MS

88

and (4) availability as neat chemicals at a reasonable price. The test chemicals cover a chemical

89

space that is highly relevant for petroleum hydrocarbons, but they were not selected to represent

90

a typical petroleum product composition.

91

Experimental

92 93

Materials. 1-octanol, 1,2,4-trimethylbenzene, 1,2-dimethylnaphthalene, 1,3,5-triethylbenzene, 2,2,4,4,6,8,8-heptamethylnonane,

9,10-dihydroanthracene,


benzo(a)pyrene,

5


Page 6 of 32

94

benzo(b)fluoranthene, benzo(k)fluoranthene, bicyclohexyl, biphenyl, trans-decalin, n-decane,

95

decylbenzene, isopentylbenzene, naphthalene, phenanthrene, pyrene, tetralin and p-xylene were

96

purchased from Sigma-Aldrich (Copenhagen, Denmark). 1,2,3,10b-tetrahydrofluoranthene,

97

1,2,3,4,5,6,7,8-octahydrophenanthrene, 1,2,3,6,7,8-hexahydropyrene, 1,2-dihydronaphthalene,

98

cis-1,2-dimethylcyclohexane,

99

diisopropylnaphthalene,

1,3,5-trimethylcyclohexane,

2,3-dimethylheptane,

2-methyl-1H-cyclopenta(l)phenanthrene,

2,6-

2-methylnonane,

4-

100

methyldodecane, benzo(a)anthracene, chrysene, m-cymene, decylcyclohexane, n-dodecane,

101

dodecylbenzene, fluoranthene, n-octylcyclohexane and p-terphenyl were purchased from TCI

102

chemicals

103

tetramethylnaphthalene,

104

ethylanthracene, 3-ethylnonane, 5α(H)-androstane, butyldecalin, dehydroabietine (96.7%),

105

fichtelite, methylenephenanthrene, perhydrofluorene and perhydropyrene were purchased from

106

Chiron (Oslo, Norway,). 5,6-dimethyl-1-(4methylpentyl)naphthalene were purchased from

107

Chemsampco via Sigma-Aldrich (Copenhagen, Denmark). Purity of chemicals was at least 97%

108

unless otherwise stated. Translucent silicone rods (custom-made by Altecweb.com, product code

109

136-8380), ethyl acetate (Sigma-Aldrich, ≥ 99.7%), ethanol (VWR chemicals, 99.8%) and

110

methanol (Sigma-Aldrich, > 99.9%) were used for passive dosing.

(Zwijndrecht,

Belgium).

1,1,4,4,6-pentamethyldecalin

2,2,5,7-tetramethyltetralin,

(91.2%),

2,6,10-trimethyldodecane

1,4,6,7-

(96.9%),

2-

111

Hydrocarbon mixtures. The 53 hydrocarbons were divided into two groups that were tested

112

in two separate biodegradation experiments. The chemicals had molecular weights between 106

113

and 252 g/mol and covered five orders of magnitude with regards to water solubility and octanol-

114

water partition ratio (Kow) and nine orders of magnitude regarding air-water partitioning (Kaw)

115

(Figure 1). Mixture 1 consisted of 35 hydrocarbons, mostly liquids. The experimental setup was

116

designed for initial concentrations of these chemicals, roughly 1000 times below their water


6

Page 7 of 32


117

solubility (see supporting information S1 for initial concentrations). Mixture 2 consisted of 19

118

solid chemicals (naphthalene was included in both mixtures to compare the two batches of water

119

used for the liquid and solid test). Initial test concentrations were approximately 10 times below

120

solubility for half of these chemicals (lowest solubility), and approximately 400 times below

121

solubility for the remaining chemicals (see S1). Test concentrations ranged from 0.004 to 170

122

µg/L. 10 4

Kaw (-)

10 2 10 0 10 -2 10 -4 10 -6 2

4

6 logK ow

Paraffins Aromatics CN

123

6-8 9-11 12-14 15-17 18-20

8

Napthenics Naphthenic Aromatics

nP iP MN DN PN MA NMA DA NDA TA NTA PA 1 1

3 1 2

1 1 1 1

1 2 1

1 1 2

1 3 1 1 1

2 2 1

1* 1+2 1 1

1 2

1 1 1

2

1 6

124

Figure 1. The chemical space of hydrocarbons included in this study regarding octanol-water

125

partition ratio, logKow25, and air-water partitioning, Kaw25, (top) and number of chemicals

126

included in each hydrocarbon block by carbon number (CN) and structural class (bottom).

127

Closed symbols denote liquid mix, open symbols indicate solid mix. nP = n-paraffin, iP = i-

128

paraffin, MN = mononaphthenic, DN = dinaphthenic, PN = polynaphthenic, MA =

129

monoaromatic, NMA = naphthenic monoaromatic, DA = diaromatic, NDA = naphthenic

130

diaromatic, TA = triaromatic, NTA = napththenic triaromatic, PA= Polyaromatic, Mixture 2 is

131

indicated in italics, *Naphthalene was included in both mixtures.


7


Page 8 of 32

132

Passive Dosing. Passive dosing can produce well defined low concentrations of hydrophobic

133

organic chemicals in water without addition of co-solvent.21,22,26 It was in the present study used

134

to set initial concentrations and mixture composition, but not to buffer concentrations during the

135

biodegradation experiment as done in previous studies.27–29

136

Silicone rods were used as passive dosing donor.23 The rods (3 mm diameter) were washed in

137

the dishwasher without soap and dried using lens cleaning tissue. 20.0 g (approx. length 2.6

138

meters) were then cut and added to 100 mL amber Wheaton glass serum bottles with crimp seals

139

and PTFE-coated silicone septa. The rods were further cleaned by soaking in ethyl acetate for

140

>20 hours, and in ethanol for > 20 hours. Ethanol was poured out and the bottles heated to 120

141

˚C for 2 hours to evaporate the remaining ethanol from the silicone.

142 143

Three different loading methods were required in order to cover the large chemical space of the present study:

144

Loading of liquid substances by full absorption. Three passive dosing systems containing

145

20.0 g silicone in 100 mL bottles were prepared. An equal mass of 35 hydrocarbons was mixed,

146

dissolving the two solid hydrocarbons (naphthalene and biphenyl) in the liquid hydrocarbons.

147

400 µL of this mixture was added to each rod. The bottles were rolled for five days at ~ 40 rpm,

148

after which the majority of the liquid mixture was visually confirmed to be absorbed into the

149

silicone, 65 mL of ultrapure water was added and the bottles rolled for 20 hours after which the

150

water was discarded. This procedure enabled the full absorption of the hydrocarbons in the

151

silicone.

152

Loading of solid hydrocarbons by partitioning from saturated methanol solutions. One

153

passive dosing system of 20.0 g silicone in a 100 mL bottle was prepared. Excess amounts of the

154

least

soluble

hydrocarbons

(pyrene,

1,2,3,6,7,8-hexahydropyrene,


p-terphenyl,

2-

8

Page 9 of 32


155

ethylanthracene,

2-methyl-1H-cyclopenta(l)phenanthrene,

benzo(a)anthracene,

chrysene,

156

benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene) were added to 25 mL methanol,

157

shaken and left overnight for equilibration and settling of crystals. In four steps, 15 mL of the

158

saturated methanol was transferred to the silicone, the bottle was rolled for >20 hours, the

159

methanol poured back to the crystals, shaken and left for settling >4 hours before being added

160

back to the silicone.

161

Loading of solid chemicals by partitioning from non-saturated methanol solution. One

162

passive dosing system of 4.0 g silicone rod placed in a 20 mL headspace vial was prepared. The

163

loading solution was prepared by adding excess amounts of the solid hydrocarbons

164

(Naphthalene,

165

dihydroanthracene, 1,2,3,10b-tetrahydrofluoranthene, phenanthrene, methylenephenanthrene,

166

2,6-diisopropylnaphthalene and fluoranthene) to 10 mL methanol, shaking and leaving overnight

167

for equilibration and settling of crystals. 5 mL of this solution was then diluted to 25 mL in

168

methanol. In three steps, approximately 8 mL of this methanol solution was added to the silicone,

169

rolled >20 hours and discarded.

1,2-dimethylnaphthalene,

1,4,6,7-tetramethylnaphthalene,

9,10-

170

The two passive dosing systems for solids were cleaned by washing 10 times with pure water

171

(1 min vigorous shaking) to remove methanol and any possible crystals from the loaded silicone

172

rods. The last fill was rolled overnight and discarded.

173

Aqueous stock solution was then prepared by adding 65 mL of ultrapure water to the 100 mL

174

passive dosing systems and 13 mL of ultrapure water to the 20 mL passive dosing system and

175

rolling at ~ 40 rpm for > 30 minutes. Stock solution was transferred to test systems using gas

176

tight syringes.


9


Page 10 of 32

177

Environmental inocula. The inocula for the biodegradation experiments originated from three

178

types of surface water: wastewater treatment plant activated sludge filtrate, seawater and lake

179

water.

180

A sample of activated sludge was taken at Lynetten wastewater treatment plant (WWTP) in

181

Copenhagen (Denmark) on the 19th of April 2016 (for liquid hydrocarbon test) and 4th of October

182

2016 (for solid hydrocarbon test). Lynetten is the largest WWTP in Denmark. The samples were

183

filtered through Whatman 114V filters (retention 20 µm) to prepare the activated sludge filtrate.

184

A surface seawater sample was taken in the North Sea west of Esbjerg, Denmark, on the 18th

185

of May 2016 (ETRS89 UTM32N: 451415; 6143811) (for liquid hydrocarbon test) and 27th of

186

September 2016 (ETRS89 UTM32N: 449694; 6142556) (for solid hydrocarbon test) in the open

187

sea, 5-8 km from the coast, 1-2 km from the main sailing route.

188

A surface sample of lake water was taken from Maglesø Lake (Sealand, Denmark), on the 5th

189

of April 2016 (for liquid hydrocarbon test) and 5th of October 2016 (for solid hydrocarbon test),

190

6-7 m from the shore. This is a clean lake, with a very small catchment, no runoff discharges and

191

situated in rural surroundings (forest/fields). This lake had the slowest biodegradation for nine

192

hydrocarbons in an initial study including two lake and three stream samples,24 and it was

193

therefore selected as the most conservative choice of inoculum in the present study.

194

All water samples were used within 24 hours of sampling. Background characterization of the

195

samples is included in supporting information (S2). The samples showed quite similar

196

characteristics at the same site between the two sampling dates, although temperatures were

197

higher in the fall than in the spring. The activated sludge filtrate sample taken in the fall showed

198

a higher degree of treatment (lower nutrients and dissolved organic carbon) than the sample

199

taken in the spring. Culturable bacterial density, measured by the heterotrophic plate count, HPC,


10

Page 11 of 32


200

was similar in the lake and seawater (approximately 1-2·103 CFU/mL) and two orders of

201

magnitude lower than in the activated sludge filtrates (2.5·105 and 4.8·105 CFU/mL).

202

Biodegradation tests. Biodegradation and final chemical analysis both took place in auto

203

sampler vials, maximizing the number of replicate test systems and minimizing test substance

204

losses between the experiment and the analysis. A large number of biotic and abiotic test systems

205

were prepared by combining 13.5 mL of environmental water containing the inoculum spiked

206

with 30 µg/L of 1-octanol (positive control substance), with 1.5 mL of aqueous stock solution

207

containing the test chemicals in 20 mL amber glass vials with screw caps and PTFE-coated

208

silicone septa. For the solid chemicals, 1.3 mL stock solution was added from the 100 mL

209

passive dosing system and 0.2 mL from the 20 mL passive dosing system. Abiotic controls were

210

prepared using ultrapure water instead of environmental samples. For liquid chemicals in

211

seawater, abiotic controls of seawater salinity were prepared by adding 35 g/L of NaCl to

212

ultrapure water. For the solid chemicals, ultrapure water was used without adding salts. The vials

213

were closed immediately and incubated at 20 °C on a bench top laboratory roller at ~ 30 rpm. At

214

time points from 2 hours to ~100 days three biotic and three abiotic test systems were taken for

215

chemical analysis (destructive sampling). While no systematic differences were seen between

216

test vials prepared from the three replicate silicone rods loaded with liquids, a balanced test

217

design was chosen where one biotic and abiotic test system were analyzed from each rod at each

218

time point. Test durations were 77 or 85 days for the activated sludge test, 56 or 111 days for the

219

seawater and 98 or 104 days for the lake water with the liquid and solid mixtures respectively.

220

Chemical analysis. Automated Solid Phase Microextraction (SPME) coupled to Gas

221

Chromatography – Mass Spectrometry (GC-MS) (Agilent Technologies 7890B/5877A

222

GC/MSD) was applied directly on the test systems. A CTC PAL RSI 85 autosampler (CTC


11


Page 12 of 32

223

Analytics, Zwingen, Switzerland) was used for SPME sampling and subsequent thermal

224

desorption of the SPME fiber. Headspace-SPME with 100 µm Polydimethylsiloxane (PDMS)

225

coating was used for the liquid chemicals, and water phase SPME with 7 µm PDMS coating was

226

used for the solid chemicals. Analytical details are included in supporting information (S3).

227

Quality assurance. Blank test systems were co-incubated and measured with each sampling

228

point. For 10 of the 2246 data points, elevated blank responses for a chemical resulted in

229

exclusion from the dataset. Two vials (1 of 246 biotic and 1 of 246 abiotic test systems) showed

230

signs of being leaky (selective loss of volatile chemicals compared to non-volatile), and were

231

excluded.

232

Background concentrations of the test chemicals in the environmental media were evaluated by

233

comparing triplicate measurements of each environmental media with the abiotic controls from

234

the same GC-MS run. In the lake and seawater background concentrations were < 1% of the test

235

concentrations. In the activated sludge filtrate, slightly higher background response levels were

236

observed, and levels above 10% were observed for 2,6,10-trimethyldodecane (15%),

237

2,2,4,4,6,8,8-heptamethylnonane

238

Diisopropylnaphthalene was therefore excluded from the data for activated sludge filtrate.

(18%)

and

2,6-diisopropylnaphthalene

(135%).

2,6-

239

In four cases, the GC peak areas of the biotic test systems were more than twice as high as for

240

the abiotic controls. The data series for 2,2,4,4,6,8,8-heptamethylnonane and 1,2,3,4,5,6,7,8-

241

octahydrophenanthrene in activated sludge filtrate as well as 5-alpha(H)-androstane and

242

Fichtelite in lake water were therefore removed from the dataset.

243

Data analysis. For each time point, the relative concentration, Crelative, was calculated as the

244

ratio between the peak area in the biotic test systems, Abiotic, relative to the peak area in the

245

abiotic test systems, Aabiotic. Since data from the abiotic and biotic test systems were not paired,


12

Page 13 of 32


246

and there is uncertainty associated with both the abiotic test systems and the biotic test systems,

247

the uncertainty related to the ratio between these two variables were calculated using Taylor

248

series approximations of the mean, µCrelative, and variance σ2Crelative, of a ratio according to

249

equation 1 and 2. 30

250

(1)

251

(2)

252

Where µAabiotic and µABiotic is the mean of the area of response in the abiotic and biotic test

253

systems, σ2Aabiotic and σ2ABiotic is the variance of the area of response in the abiotic and biotic test

254

systems, and Cov(ABiotic,Aabiotic) is the co-variance between the area of response in the biotic and

255

abiotic test systems.

256

The mean and uncertainty of Crelative for each time point was used as input to GraphPad Prism

257

5.00 for fitting the first order degradation model with lag phase in equation 3. Tlag was

258

constrained to positive values, no weighting of the data was used, and the number and scatter

259

among replicates was accounted for in the fit. Confidence limits for ksystem were obtained

260

assuming lognormal distribution of ksystem.

261

(3)

262

Test system half-lives, T½,system, were obtained by equation 4.

263

(4)


13


264 265

Page 14 of 32

Degradation half-times in the system, DT50, were obtained by summing the lag-phase and halflife.

266

The biodegradation experiments were performed with a headspace in the test systems to ensure

267

aerobic conditions throughout the incubation (necessary for the activated sludge filtrate). Test

268

system half-lives of the volatile chemicals were thus corrected for headspace partitioning in

269

order to obtain water phase half-lives (T½,water) as described by Birch et al.20 This correction did

270

not account for the hydrocarbon binding to third phases such as dissolved organic matter.

271

When degradation proceeds, the concentrations in biotic test systems will at some point reach a

272

level below detection. This is important information, but the low concentration measurements

273

should not be given too much weight in the fitting of the degradation curve. In the present study,

274

three times the peak area of the blank response was used as the limit of detection, and for each

275

degradation curve, a maximum of one data point was included below this limit. Furthermore, a

276

maximum of three data points of Crelative < 0.01, were included in each time series.

277 278

The degradation curves were evaluated based on the coefficient of determination of the fit (R2) and visual inspection, and were subsequently divided into six categories:

279

(1.1) High quality model fit with a goodness of fit of R2>0.8. For this category the degradation

280

rate constant, lag-phase, half-life and half-times were determined.

281

(1.2) Goodness of fit was R2 test duration).

292

(3.2) Limited degradation. One time point had Crelative < 0.5. For this category, the time of the

293

last data point before the data point below 0.5 was used to assign a minimum half-time.

294

The threshold of 50% reduction relative to abiotic controls for positive identification of

295

degradation in category 2, 3.1 and 3.2 was chosen as a common criterion for all chemicals and

296

degradation curves. It was based on typical variability between replicates at each data point seen

297

for chemicals with slow degradation.

298

Results and Discussion

299

Biodegradation kinetics. Degradation curves and the associated biodegradation kinetic

300

parameters such as lag phase, first order rate constant, half-times, test system half-lives and

301

water phase half-lives are listed in Supporting Information (S4 and S5). 1-octanol was degraded

302

within 2 days in both batches for all three water types, confirming the biological activity of all

303

samples. Examples of degradation curves fitted to experimental data are shown for five

304

hydrocarbons in Figure 2. The three iso-paraffins in Figure 2 are examples of a category 1.1 fit in

305

all three water types, 1,2-dimethylnaphthalene is an example of a category 1.2 fit in the activated

306

sludge filtrate, and 1.3 fit in the seawater and lake water. 1,4,6,7-Tetramethylnaphthalene is a

307

category 1.1 fit in the activated sludge filtrate, a category 2 fit in the seawater, and a category 3.1

308

fit in the lake water. The number of hydrocarbons within each of the degradation curve

309

categories is shown in Supporting Information (S6).


15


Crelative

Activated sludge filtrate 1.5

1.5

1.0

1.0

1.0

0.5

0.5

0.5

0

20

40

60

80

0

100

1.5

1.5

1.5

1.0

1.0

1.0

0.5

0.5

0.5

0.0

2,3-Dimethylheptane 2-Methylnonane 3-Ethylnonane

0.0

0.0 0 5 10 15 20

Crelative

Lake water

Seawater

1.5

0.0

Page 16 of 32

0.0 0 5 10 15 20 25 0

Incubation time (days)

20

40

60

80

100

1,2-dimethylnaphthalene 1,4,6,7-tetramethylnaphthalene

0.0 20

40

60

80

100

0


20

40

60

80

100


310 311 312

Figure 2: Relative concentration for three C9-C11 iso-paraffins (top) and for two C12-C14

313

diaromatics (bottom) and first order degradation curves. 1,4,6,7-tetramethylnaphthalene in

314

seawater is shown in gray open symbols since it was discarded based on inconsistent data

315

(category 2). Error bars show standard error of mean (n=3) based on three replicate biotic and

316

three replicate abiotic test systems for each time point.

317

Biodegradation differences between water types. Of 53 hydrocarbons, first order

318

degradation kinetics (category 1.1 + 1.2) were obtained for 40 hydrocarbons in the activated

319

sludge filtrate, 18 hydrocarbons in the seawater and 21 hydrocarbons in the lake water. Half-lives

320

in activated sludge were shorter than in lake water for all hydrocarbons and shorter than in

321

seawater for most hydrocarbons. A slightly lower number of hydrocarbons were degraded in the

322

seawater compared to the lake water, however, when degradation initiated, half-lives were with

323

few exceptions similar or slightly shorter in the seawater. While the heterotrophic plate counts

324

were similar in the sea- and lake water samples, the seawater samples were taken in the vicinity

325

of a trafficked shipping port, which likely implied pre-exposure of the natural bacterial consortia


16

Page 17 of 32


326

to petroleum hydrocarbons. The lake was to the contrary located in a rural area without runoff

327

discharges.

328

Inoculum from the lake was in a previous study found to give the slowest hydrocarbon

329

biodegradation among two lakes and three streams, and seven out of nine hydrocarbons were

330

degraded even slower in the present study compared to the previous study (e.g. DT50 for trans-

331

decalin was >56 days in this study while 35 days in the previous study).24 The aromatic

332

chemicals in the liquid mixture test in lake water showed highly varying results with no clear

333

first order degradation curves, and Naphthalene was only degraded in lake water in the

334

experiment including the solid hydrocarbons. Although theoretically, the difference in

335

degradation of naphthalene between the two batches of lake water could be caused by the

336

difference in the constituents of the two mixtures, we hypothesize that these differences and

337

inconsistencies in biodegradation kinetics for the lake water in this experiment was caused by an

338

insufficient number of competent aromatic degraders in the 15 mL test systems prepared from

339

the sample taken on the first sampling date. This explanation is in line with a recent study

340

showing that higher inoculum concentrations in screening tests can increase reproducibility of

341

results because it reduces the risk of excluding specific degraders in the test volume.31

342

Hydrocarbon groups. Figure 3 shows the range of DT50 for the tested chemicals in each

343

hydrocarbon group in activated sludge filtrate, seawater and lake water. Note that there are a

344

different number of chemicals tested within the different groups (see Figure 1), and for some

345

chemicals (category 2) a DT50 was not determined as described above.


17


Naphthenics

Paraffins

n-P

DN

10

DT 50 (days)

10 0

Aromatics

1

MN 0. 1

DT 50 (days)

10 0

10

1

PolyN

NMAr DT 50 (days)

Activated sludge filtrate

Seawater

DT 50 (days)

10 0

MAr

0. 1

NDiAr

10 0

DiAr

10

NTriAr

1

TriAr

10

Naphthenic Aromatic

PolyAr

1

0. 1

i-P

0. 1

Page 18 of 32

Lake water

346 347

Figure 3: Half-times for hydrocarbon groups in activated sludge filtrate, seawater and lake

348

water. Median, 25th, 75th percentiles, whiskers show minimum to maximum range. For

349

hydrocarbon group abbreviations, see Figure 1.

350

The linear paraffins in this test were degraded faster than the branched paraffins. For aromatic

351

hydrocarbons a trend was seen of longer half-lives with increasing number of rings. The

352

sequence of degradation based on median DT50 for each group is in agreement with earlier

353

observations of hydrocarbon group susceptibility to biodegradation (n-paraffin < i-paraffin < low

354

molecular weight aromatic < napththenic / high molecular weight aromatic).32–34 Large

355

differences in half-times were observed within some of these groups however, and the span of

356

degradation half-times overlapped for all hydrocarbon groups.

357

Chemical space. DT50 and T½,water for the test chemicals in seawater were grouped into ranges

358

and plotted within the chemical space in terms of their air-water (Kaw) and octanol-water


18

Page 19 of 32


359

(logKow) partition ratio (Figure 4). Ranges were chosen to represent very fast degradation (40 days).

Kaw (-)

10 4

A

DT50 40 d

10 -2

No DT50

10 -4 10 -6 2 10 4

Kaw (-)

10

4

6 logKow

8

B

T½,w ater 40d) covered 3 orders of magnitude

382

in logKow and 5 orders of magnitude in Kaw, without a clear grouping in any region of the two-

383

dimensional space shown in Figure 4 (orange triangles). While the two physicochemical

384

properties were useful to describe the chemical space of the tested hydrocarbons, they seem

385

poorly related to the biodegradation endpoints of the study. This observation is specific to

386

biodegradation testing in water, since increases in sorption with increasing hydrophobicity can

387

induce strong relationships between hydrophobicity and biodegradation half-lives in soils and

388

sediments.35 Whereas sorption may reduce the bioavailable fraction, and thus reduce

389

biodegradation rates, when sediments are included in tests, sediments have also been observed to

390

increase biodegradation rates because of the increase in sediment associated bacteria.36,37 The

391

most persistent hydrocarbons were also distributed widely between hydrocarbon classes and

392

carbon number groups, again without a clear trend and grouping. While the carbon block

393

approach again is very useful to describe the chemical space of hydrocarbons, there are clearly

394

additional structural features beyond the groups, which determine the biodegradability. For

395

example, a structural factor that resulted in fast primary degradation in all waters was the

396

inclusion of a long (>C4) linear alkyl chain. Higher methyl substitution resulted in slower


20

Page 21 of 32


397

degradation for naphthalene, 1,2-dimethylnaphthalene and 1,4,6,7-tetramethylnaphtalene, and

398

aromatic ring structures such as naphthalene and pyrene were degraded faster than their

399

naphthenic analogues in the activated sludge filtrate, and in some cases in the lake- and seawater.

400

Comparing the experimental data to the BioHCwin model. A comparison between

401

BioHCwin predicted half-lives and T½,water or DT50 from this study is shown in Figure 5. Both

402

end-points (first order half-lives and degradation half-times) are relevant in an environmental

403

context. Half-lives are relevant for biodegradation of diffuse on-going emissions and is the end-

404

point used from simulation tests (such as OECD 309) to compare to persistency criteria.

405

Degradation half-times are more relevant to spill scenarios and are used in screening studies.

Predicted half-life (days)

Predicted half-life (days)

Lake water

Seawater


1000

1000

1000

100

100

100

10

10

10

1 0.1

1

10

100

1 0.1

1

10

100

1 0.1

10

T ½,water (days)

T ½,wate r (days)

Lake water

Seawater


1000

1000

1000

100

100

100

10

10

10

1 0.1

1

T ½,water (days)

1

10

100

1 0.1

1

10

100

DT 50 (days)

DT 50 (days) Paraffins

Aromatics

Naphthenics

1 0.1

1

10

100

100

DT 50 (days) NaphthenicAromatics

406 407

Figure 5: BioHCwin predicted half-lives vs experimental water phase half-lives (top row) in

408

lake water, seawater and activated sludge filtrate. BioHCwin predicted half-lives vs test system

409

half-times (bottom row). Open symbols indicate minimum half-times of hydrocarbons with


21


Page 22 of 32

410

limited degradation during the test. 1:1 is indicated as a solid line and 10 times under and over-

411

prediction as dotted lines.

412

Since biodegradation is not an inherent property of the chemical, and environmental factors

413

such as sorption and microbial activity are important for biodegradation rates, half-lives can

414

easily vary a factor 10 between studies with different environmental conditions.6 Even within

415

batches of water from the same site, variations occur, as seen for Naphthalene in the two

416

mixtures. The BioHCwin model was developed to predict degradation in “different

417

environmental media (e.g., water, soil, and sediment)”,6 and more than half of the input data was

418

from sediment and soil studies. It is unclear whether the included studies were detailed enough to

419

determine the lag-phase and report true first order degradation rates. Our T½,water data set targets

420

degradation in the water phase only and is thus not necessarily directly comparable to the

421

BioHCwin model.

422

Activated sludge filtrate has a higher bacterial density and contains better adapted bacteria than

423

surface water and seawater, and the use of this type of data was limited in the BioHCwin model

424

development. BioHCwin predictions for environmental half-lives were therefore similar or

425

longer than all experimental water phase half-lives (T½,water) in activated sludge filtrate, and

426

higher or within a factor 10 of the DT50s.

427

Generally the predicted half-lives were within a factor 10 or higher than T½,water and DT50 in

428

sea- and lake water (Figure 5). The higher half-lives could be explained by the inclusion of

429

sediment and soil studies in the training set for the BioHCwin model or if the bacteria in the

430

current seawater sample was more pre-adapted to petroleum hydrocarbons than the inoculum in

431

studies used to develop the BioHCwin model. A study by Prosser et al.16 also reported higher

432

half-life predictions by BioHCwin compared to biodegradation data summarized from literature


22

Page 23 of 32


433

in seawater and stormwater pond water in which there were no sediment or soil present. Three

434

additional explanations for the generally faster degradation in the present study compared to

435

BioHCwin predictions are: (1) Biodegradation testing in gas tight vials allowed the

436

determination of water phase first order half-lives.20 (2) Biodegradation testing at lower and

437

more environmentally relevant substrate concentrations can lead to higher biodegradation rate

438

constants.27 (3) The use of passive dosing for setting initial concentrations circumvents the

439

testing of dispersed micro droplets of pure hydrocarbons. This ensured that biodegradation

440

kinetics did not become rate limited by the dissolution of the non-dissolved phase. A comparison

441

between the seawater half-lives in the present study and calculated well adapted in situ half-lives

442

from the Deepwater Horizon oil spill17, showed similar half-lives for 1,2,4-trimethylbenzene (1.8

443

days in the present study and 0.8 days in situ) whereas phenanthrene had a longer half-life in this

444

study (16 days) compared to the in situ calculated half-life (1.6 days).

445

A detailed look at the results showed that BioHCwin highly overpredicted the half-lives of a

446

few chemicals: Dehydroabietine and 1,2,3,10b-tetrahydrofluoranthene, for example, had

447

predicted half-lives of 2819 and 4908 days and observed half-times of 12 and 63 days in lake

448

water respectively. These predicted half-lives are not reliable as they are outside of the input data

449

range for the model calibration, and are probably resulting from a lack of data for naphthenic di-

450

and triaromatic hydrocarbons.6,16

451

Underprediction of half-lives is more problematic than overprediction since it can lead to risks

452

being

overlooked.

The

two

cycloalkanes

cis1,2-dimethylcyclohexane

453

trimethylcyclohexane had BioHCwin predicted half-lives of 5.1 and 3.5 days but were not

454

degraded in the seawater nor in the lake water. In activated sludge filtrate they were degraded

455

with T½,water (1.4 and 4.5 days) close to the BioHCwin predictions. Lack of or slow degradation


and

1,3,5-

23


Page 24 of 32

456

was similarly seen for 1,3,5-trimethylcyclohexane in a number of surface water samples in

457

earlier similar experiments,24 but faster degradation was observed in other studies in seawater

458

and rainwater retention pond water using higher initial chemical concentrations.16 Two other

459

mono-naphthenic structures included here had long linear alkyl chains, and primary degradation

460

was presumably driven by this chain rather than their naphthenic structure. It is however

461

noteworthy that the di- and tri- naphthenic structures decalin, bicyclohexyl and

462

perhydrofluorene, which did not include any alkyl chains, were degraded in the seawater.

463

The most persistent of the hydrocarbons included in this study were 1,1,4,4,6-

464

pentamethyldecalin, perhydropyrene and 1,2,3,6,7,8-hexahydropyrene, which had limited

465

degradation in all three types of water. Furthermore, 2,2,4,4,6,8,8-heptamethylnonane showed

466

inconsistent degradation in the activated sludge filtrate and seawater and did not degrade in the

467

lake water. In line with observations from Comber et al.,15 these hydrocarbons were either highly

468

branched with quarternary carbons or highly cyclic structures, and had β-substituted terminal

469

carbons (see S7), which prevents β-oxidation as a primary transformation step.38 BioHCwin

470

identified these structures as slow to degrade with half-life predictions of 129 - 451 days, which

471

is longer than the test duration in this study.

472

The application of passive dosing for setting initial hydrocarbon concentrations in

473

combination with the very close alignment of test system and SPME-GC-MS analysis provided

474

new possibilities for biodegradation testing. The chosen test volume of 15 mL, was a

475

compromise between the 100-1000 mL test systems often used in regulatory biodegradation

476

studies3,39 and high-throughput miniaturized systems40,41. This reduced test volume was very

477

practical and appeared sufficient for the biodegradation testing with the WWTP sludge filtrate

478

and seawater. However, for the lake water with limited biodegradation activity we observed


24

Page 25 of 32


479

larger deviations between replicate test systems and a higher frequency of inconsistent data,

480

which might indicate an insufficient test volume for such samples and asks for further studies.

481

The parallel biodegradation testing of up to 34 test chemicals was shown to be a very time and

482

cost efficient approach for the generation of a large and consistent data set of biodegradation

483

kinetic data while minimizing the effect of confounding factors. However, further research is

484

needed for determining possible co-substrate effects on the biodegradation kinetics at low

485

concentrations. The use of pure water as abiotic control instead of poisoned controls was

486

considered appropriate for these aquatic tests with surface water or activate sludge filtrate, but

487

might require adjustments when increasing the amount of suspended particles. In case of

488

concern, poisoned controls can be used instead. Overall the new approach had several

489

advantages, but of course also limitations. The advantages of this approach are mainly (1) the

490

potential to generate large data sets for chemicals covering a large and relevant chemical space,

491

(2) the possibility to conduct degradation studies at very low environmentally relevant

492

concentrations while avoiding dispersions of pure phase and (3) the minimization of

493

experimental steps which facilitates biodegradation testing with native microorganisms and

494

rather volatile test substances. The main limitations of the approach are that it is based on

495

substrate depletion, which limits it to the study and quantification of primary biodegradation and

496

that it presently is limited to biodegradation testing in aqueous media.

497

ASSOCIATED CONTENT

498

Supporting Information. Initial concentrations, characterization of environmental samples,

499

details of GC-MS analysis, kinetics, degradation curves and structure of chemicals with limited

500

degradation in all waters. This material is available free of charge via the Internet at

501

http://pubs.acs.org.


25


Page 26 of 32

502

AUTHOR INFORMATION

503

Corresponding Author

504

* Technical University of Denmark, Department of Environmental Engineering, Bygningstorvet,

505

Building 115, 2800 Kgs. Lyngby, Denmark. [email protected].

506

ACKNOWLEDGMENT

507

The authors thank Concawe for financial support, Chris Hughes, Mike Comber, Thomas

508

Parkerton, Aaron Redman and the Concawe Ecology Group for comments on the draft

509

manuscript, Hanne Bøggild for technical assistance in the laboratory and Lynetten wastewater

510

treatment plant for providing activated sludge.

511

REFERENCES

512 513

(1)

European Parliament. Regulation (EC) No 1907/2006 of the European Parliament and of

the Council of 18 December 2006.

514

(2)

515

2002.

516

(3)

517 518

US Public Law. Toxic Substances Control Act; U.S. Senate: 15 US Code, Chapter 53,

OECD 301. OECD Guideline for Testing of Chemicals. Ready Biodegradability. 1992,

pp 1–62. (4)

Aronson, D.; Boethling, R.; Howard, P.; Stiteler, W. Estimating biodegradation half-lives

519

for

520

10.1016/j.chemosphere.2005.09.044

521 522

(5)

use

in

chemical

screening.

Chemosphere

2006,

63

(11),

1953–1960.

DOI

Rücker, C.; Kümmerer, K. Modeling and predicting aquatic aerobic biodegradation – a

review from a user’s perspective. Green Chem. 2012, 14 (0), 875. DOI 10.1039/c2gc16267a


26

Page 27 of 32

523


(6)

Howard, P.; Meylan, W.; Aronson, D.; Stiteler, W.; Tunkel, J.; Comber, M.; Parkerton,

524

T. F. A new biodegradation prediction model specific to petroleum hydrocarbons. Environ.

525

Toxicol. Chem. 2005, 24 (8), 1847–1860. DOI 10.1897/04-453R.1

526

(7)

Rorije, E.; Verbruggen, E. M. J.; De Knecht, J. A. Framework Contract No ECHA /

527

2008. Service Request on a critical review of the environmental and physicochemical

528

methodologies commonly employed in the environmental risk assessment of petroleum

529

substances

530

http://echa.europa.eu/documents/10162/13628/review_environmental_physicochemical_method

531

ol_en.pdf

532 533 534

(8)

in

the

context

of

REACH

registrations.

2012.

Quann, R. J. Modeling the chemistry of complex petroleum mixtures. Environ. Health

Perspect. 1998, 106, 1441–1448. DOI 10.1289/ehp.98106s61441 (9)

van de Meent, D.; Hollander, A.; Comber, M.; Parkerton, T. Environmental Fate Factors

535

and Human Intake Fractions for Risk Assessment of Petroleum Products. Integr. Environ.

536

Assess. Manag. 2009, 6 (1), 135–144. DOI 10.1897/IEAM_2009-035.1

537

(10) King, D. J.; Lyne, R. L.; Girling, A.; Peterson, D. R.; Stephenson, R.; Short, D.

538

Environmental risk assessment of petroleum substances : the hydrocarbon block method;

539

Concawe: Brussels, Report No. 52/96, 1996; Vol. 52.

540 541 542 543

(11) Prince, R. C.; Parkerton, T.; Lee, C. The primary aerobic biodegradation of gasoline hydrocarbons. Environ. Sci. Technol. 2007, 41, 3316–3321. DOI 10.1021/es062884d (12) Prince, R. C.; Haitmanek, C.; Lee, C. C. The primary aerobic biodegradation of biodiesel B20. Chemosphere 2008, 71 (8), 1446–1451. DOI 10.1016/j.chemosphere.2007.12.010


27


Page 28 of 32

544

(13) Prince, R. C.; McFarlin, K. M.; Butler, J. D.; Febbo, E. J.; Wang, F. C. Y.; Nedwed, T. J.

545

The primary biodegradation of dispersed crude oil in the sea. Chemosphere 2013, 90 (2), 521–

546

526. DOI 10.1016/j.chemosphere.2012.08.020

547

(14) Brakstad, O. G.; Nordtug, T.; Throne-Holst, M. Biodegradation of dispersed Macondo oil

548

in seawater at low temperature and different oil droplet sizes. Mar. Pollut. Bull. 2015, 93 (1–2),

549

144–152. DOI 10.1016/j.marpolbul.2015.02.006

550

(15) Comber, M. I. H.; den Haan, K. H.; Djemel, N.; Eadsforth, C. V.; King, D.; León

551

Paumen, M.; Parkerton, T.; Dmytrasz, B. Primary Biodegradation of Petroleum Hydrocarbons

552

in Seawater; Concawe: Brussels, Report No. 10/12, 2012.

553

(16) Prosser, C. M.; Redman, A. D.; Prince, R. C.; Paumen, M. L.; Letinski, D. J.; Butler, J.

554

D. Evaluating persistence of petroleum hydrocarbons in aerobic aqueous media. Chemosphere

555

2016, 155, 542–549. DOI 10.1016/j.chemosphere.2016.04.089

556

(17) Thessen, A. E.; North, E. W. Calculating in situ degradation rates of hydrocarbon

557

compounds in deep waters of the Gulf of Mexico. Mar. Pollut. Bull. 2017, 122 (1–2), 77–84.

558

DOI 10.1016/j.marpolbul.2017.06.004

559 560

(18) Prince, R. C.; Butler, J. D.; Redman, A. D. The rate of crude oil biodegradation in the sea. Environ. Sci. Technol. 2017, 51 (3), 1278–1284. DOI 10.1021/acs.est.6b03207

561

(19) Lee, K.; Nedwed, T.; Prince, R. C.; Palandro, D. Lab tests on the biodegradation of

562

chemically dispersed oil should consider the rapid dilution that occurs at sea. Mar. Pollut. Bull.

563

2013, 73, 314–318. DOI 10.1016/j.marpolbul.2013.06.005


28

Page 29 of 32


564

(20) Birch, H.; Andersen, H. R.; Comber, M.; Mayer, P. Biodegradation testing of chemicals

565

with high Henry’s constants – Separating mass and effective concentration reveals higher rate

566

constants. Chemosphere 2017, 174, 716–721. 10.1016/j.chemosphere.2017.02.003

567

(21) Birch, H.; Gouliarmou, V.; Lützhoft, H. C. H.; Mikkelsen, P. S.; Mayer, P. Passive

568

dosing to determine the speciation of hydrophobic organic chemicals in aqueous samples. Anal.

569

Chem. 2010, 82 (3), 1142–1146. DOI 10.1021/ac902378w

570

(22) Gouliarmou, V.; Smith, K. E. C.; De Jonge, L. W.; Mayer, P. Measuring binding and

571

speciation of hydrophobic organic chemicals at controlled freely dissolved concentrations and

572

without phase separation. Anal. Chem. 2012, 84, 1601–1608. DOI 10.1021/ac2028497

573

(23) van der Heijden, S. a.; Hermens, J. L. M.; Sinnige, T. L.; Mayer, P.; Gilbert, D.; Jonker,

574

M. T. O. Determining High-Quality Critical Body Residues for Multiple Species and Chemicals

575

by Applying Improved Experimental Design and Data Interpretation Concepts. Environ. Sci.

576

Technol. 2015, 49, 1879–1887. DOI 10.1021/es505078r

577

(24) Birch, H.; Hammershøj, R.; Comber, M.; Mayer, P. Biodegradation of hydrocarbon

578

mixtures in surface waters at environmentally relevant levels - Effect of inoculum origin on

579

kinetics

580

10.1016/j.chemosphere.2017.05.169

and

sequence

of

degradation.

Chemosphere

2017,

184,

400–407.

DOI

581

(25) National Food Institute DTU. Danish (Q)SAR Database. National Food Institute,

582

Technical University of Denmark. http://qsardb.food.dtu.dk/database/index.html (accessed Jun

583

30, 2017).


29


Page 30 of 32

584

(26) Smith, K. E. C.; Dom, N.; Blust, R.; Mayer, P. Controlling and maintaining exposure of

585

hydrophobic organic compounds in aquatic toxicity tests by passive dosing. Aquat. Toxicol.

586

2010, 98 (1), 15–24. DOI 10.1016/j.aquatox.2010.01.007

587

(27) Smith, K. E. C.; Rein, A.; Trapp, S.; Mayer, P.; Karlson, U. G. Dynamic passive dosing

588

for studying the biotransformation of hydrophobic organic chemicals: Microbial degradation as

589

an example. Environ. Sci. Technol. 2012, 46 (9), 4852–4860. DOI 10.1021/es204050u

590

(28) Tejeda-Agredano, M. C.; Mayer, P.; Ortega-Calvo, J. J. The effect of humic acids on

591

biodegradation of polycyclic aromatic hydrocarbons depends on the exposure regime. Environ.

592

Pollut. 2014, 184, 435–442. DOI 10.1016/j.envpol.2013.09.031

593

(29) Lee, Y. S.; Lee, D. H. Y.; Delafoulhouze, M.; Otton, S. V.; Moore, M. M.; Kennedy, C.

594

J.; Gobas, F. A. P. C. In vitro biotransformation rates in fish liver S9: Effect of dosing

595

techniques. Environ. Toxicol. Chem. 2014, 33 (8), 1885–1893. DOI 10.1002/etc.2636

596 597

(30) Seltman,

H.

Approximations

for

Mean

and

Variance

of

a

Ratio

www.stat.cmu.edu/~hseltman/files/ratio.pdf (accessed Jun 29, 2017).

598

(31) Martin, T. J.; Snape, J. R.; Bartram, A.; Robson, A.; Acharya, K.; Davenport, R. J.

599

Environmentally Relevant Inoculum Concentrations Improve the Reliability of Persistent

600

Assessments in Biodegradation Screening Tests. Environ. Sci. Technol. 2017, 51 (5), 3065–3073.

601

DOI 10.1021/acs.est.6b05717

602 603

(32) Leahy, J. G.; Colwell, R. R. Microbial degradation of hydrocarbons in the environment. Microbiol. Rev. 1990, 54 (3), 305–315.


30

Page 31 of 32


604

(33) Gros, J.; Reddy, C. M.; Aeppli, C.; Nelson, R. K.; Carmichael, C. A.; Arey, J. S.

605

Resolving biodegradation patterns of persistent saturated hydrocarbons in weathered oil samples

606

from the Deepwater Horizon disaster. Environ. Sci. Technol. 2014, 48 (3), 1628–1637. DOI

607

10.1021/es4042836

608 609

(34) Head, I. M.; Jones, D. M.; Röling, W. F. M. Marine microorganisms make a meal of oil. Nat. Rev. Microbiol. 2006, 4 (3), 173–182. DOI 10.1038/nrmicro1348

610

(35) Rein, A.; Fernqvist, M. M.; Mayer, P.; Trapp, S.; Bittens, M.; Karlson, U. G. Degradation

611

of PCB congeners by bacterial strains. Appl. Microbiol. Biotechnol. 2007, 77, 469–481. DOI

612

10.1007/s00253-007-1175-6

613

(36) Shrestha, P.; Junker, T.; Fenner, K.; Hahn, S.; Honti, M.; Bakkour, R.; Diaz, C.;

614

Hennecke, D. Simulation Studies to Explore Biodegradation in Water-Sediment Systems: From

615

OECD 308 to OECD 309. Environ. Sci. Technol. 2016, 50 (13), 6856–6864. DOI

616

10.1021/acs.est.6b01095

617

(37) Xia, X. H.; Yu, H.; Yang, Z. F.; Huang, G. H. Biodegradation of polycyclic aromatic

618

hydrocarbons in the natural waters of the Yellow River: Effects of high sediment content on

619

biodegradation. Chemosphere 2006, 65 (3), 457–466. DOI 10.1016/j.chemosphere.2006.01.075

620 621 622 623

(38) Alexander, M. Nonbiodegradable and other recalcitrant molecules. Biotechnol. Bioeng. 1973, 15, 611–647. DOI 10.1002/bit.260150402 (39) OECD 309. OECD Guideline for the testing of Chemicals. Aerobic Mineralisation in Surface Water – Simulation Biodegradation Test. OECD 2004.


31


Page 32 of 32

624

(40) Cregut, M.; Jouanneau, S.; Brillet, F.; Durand, M. J.; Sweetlove, C.; Chenèble, J. C.;

625

L’Haridon, J.; Thouand, G. High throughput and miniaturised systems for biodegradability

626

assessments. Environ. Sci. Pollut. Res. 2014, 21 (16), 9545–9552. DOI 10.1007/s11356-013-

627

2236-2 (41) Martin, T. J.; Goodhead, A. K.; Acharya, K.; Head, I. M.; Snape, J. R.; Davenport, R. J.

629

High Throughput Biodegradation-Screening Test To Prioritize and Evaluate Chemical

630

Biodegradability.

631

10.1021/acs.est.7b00806

Relative Concentration

628

Sci.

Technol.

2017,

51

(12),

7236–7244.

DOI

1.0

0.5

0.0 0

632

Environ.

50

100



32

Determining biodegradation kinetics of ... - ACS Publications

Recommend Documents